ABC | Volume 113, Nº2, August 2019

Original Article Casonatto et al. Citrulline and post-exercise hypotension Arq Bras Cardiol. 2019; 113(2):218-228 Table 3 – Effect size from Paired t test (versus rest [d = mean/SD]) Control-Placebo Control-Citrulline Exercise-Placebo Exercise-Citrulline ES p ES p ES p ES p SBP 60 min 0.61 0.084 -0.56 0.109 0.53 0.127 1.81 < 0.001 Awake 0.69 0.055 0.26 0.428 1.00 0.011 0.74 0.042 Asleep 0.73 0.044 0.65 0.068 0.83 0.027 0.78 0.034 24 hours 0.92 0.017 0.48 0.156 0.57 0.102 0.88 0.020 DBP 60 min –0.12 0.717 –0.91 0.018 –0.12 0.707 –0.01 0.968 Awake 0.32 0.333 0.71 0.051 0.81 0.030 2.06 < 0.001 Asleep 0.84 0.027 1.21 0.004 0.90 0.019 2.16 < 0.001 24 hours 0.04 0.905 0.97 0.014 0.65 0.071 2.27 < 0.001 SD: standard deviation; ES: effect-size; SBP: systolic blood pressure; DBP: diastolic blood pressure. The percentages of responders and non-responders in the different experimental groups for different time periods (60 min, awake, asleep, and 24 hours) are shown in table 4. The percentages of the systolic BP responders varied from 20% (CC) to 90% (EC) for 60 min; 60% (CC and EP) to 100% (EC) for awake; 60% (CC and EP) to 90% (EC) for asleep; and 60% (CC and EP) to 90% (EC) for 24 hours. The percentages of the diastolic BP responders varied from 10% (CC) to 40% (CP) for 60 min; 50% (CP) to 100% (EC) for awake; 70% (CP) to 100% (CC and EC) for asleep; and 60% (CP) to 100% (EC) for 24 hours. Absolute systolic and diastolic BP individual changes for each experimental group in the 60 min, awake, asleep, and 24 hours are presented in Figures 1 (systolic) and 2 (diastolic). All participants in the EC demonstrated a reduction in systolic BP in the 60 min, awake, and 24 hours. For the asleep period, only one EC participant (number 23) did not present a reduction in systolic BP (Figure 2). All participants in the EC presented a reduction in diastolic BP in the awake, asleep, and 24 hours (Figure 3). Discussion The main findings of this study were: (i) there was considerable inter-individual responsiveness variability in systolic and diastolic BP responses following all experimental protocols, (ii) despite the inter-individual responsiveness variability, there was PEH only when associated with CM supplementation. To the best of our knowledge, this is the first report on inter-individual analysis of PEH following a single bout of aerobic exercise preceded by CM supplementation in hypertensive subjects. Unlike the other experimental groups, there were no “non‑responders” in the EC for “awake” (systolic and diastolic BP) and “24 hours” (diastolic BP). These results are interesting since “non-responders” have generally been identified both in non-hypertensive 9,10 and hypertensive 5,6 populations. It is important to clarify that in an intervention group, some individuals may demonstrate improvement in a given outcome (responder) whereas others show no changes or even an adverse response to exercise training (non-responders or adverse responders). 9 In fact, the heterogeneity of responses to exercise training is a current concern in exercise investigations 29 since even homogeneous samples may demonstrate a wide range of heterogeneous responses. 12 In addition, results of PEH are generally limited to presenting the average reductions in the experimental group, ignoring inter-individual variations 12 which can lead to misinterpretation since not all individuals respond favorably. It is important to consider that the results reported by central tendency measures can vary depending on the analysis. In medical studies, it is common to take measurements before and after medical interventions. How to measure the changes from baseline is a common question posed by researchers. For example, in the present study, when absolute values were used for inter-group comparisons (Table 2), no statistical differences were found for systolic BP. On the other hand, using the percentage difference (delta), a significant reduction was found in the EC for “60 min” post-exercise. Another inconsistency can be seen in diastolic BP analysis. When absolute values were used for inter-group comparisons, a significant difference was found in the EC for “24 hours” post‑exercise. On the other hand, using the percentage difference, significant reductions were found in the EC for “awake” and "24 hours" post-exercise. In specific situations, some studies (e.g., Vickers) 30 suggested avoiding using percentage changes. Despite these inconsistencies, clinicians may prefer to choose the method that will most obviously demonstrate the health‑improvement. Some researchers may choose the method that can be best understood by the majority of people interested in the research. However, one of the most common questions driving the evaluation of intervention programs is "how does this effect compare with the effects of other interventions?". Therefore, previous studies have encouraged the use of “effect size” in quantitative studies. 31 Researchers are often stimulated 222